The open source network effect depends on unrestricted software freedoms. Licensing & business models that restrict those freedoms aren’t seeking the open source effect – or if they are they will fail – so calling a policy, product or company that does so “open source” is false advertising.

A focus solely on open source legal and licensing matters as they affect companies creates bad outcomes — for leaders and their advisers who are surprised by community and market reactions, and for developers who feel abused and betrayed by “open source companies” and “government initiatives” that actually put obstacles in their path rather than remove them.  While the minutiae of open source licensing and governance need to be understood and accommodated, it’s vital to never lose sight of the open source effect itself.

The Open Source Effect

When we advocate for open source we advocate for the disruptive network effect free software produces stochastically. The open source effect everyone seeks happens when developers go ahead solving problems using the solutions of their predecessors and peers, without having to constantly ask for approval. Developers use solutions that deliver the capabilities they want with the least bureaucratic friction. Adoption then spreads organically. As they deploy, they fix bugs, report problems and improve capabilities, pooling the innovation — not least to reduce the drag of maintenance — because of the software freedoms they enjoy from the OSI-approved licensing. Great code, innovative deployers and rapid improvement catalyse exponential adoption.

There is great risk from conflating corporate legal worldviews and the worldviews of developers who make the Open Source Effect actually happen. In some cases companies are transitioning out of open source and know they must cover for their community betrayal by blaming others. In other cases, there’s market demand for open source but players dependent on royalty revenues and market control fear for their business levers and demand compromise on details. They are aware that open source leads to increased innovation and competition and they fear disruption by both.

It’s a significant European issue at the moment, with the Commission itself and important European SDOs being fed corporate legal worldviews with no context of community realities and heading into outcomes where the open source network effect simply won’t materialise as a consequence – possibly an intended one.

Woods vs Trees

Decision makers are encouraged to focus on the existing trees, and not be distracted by the current and potential woodlands. The whispered lobbying aims to validate non-OSI-approved licenses as acceptable to open source and the traditional mechanisms of royalty collection and market control as necessary to accommodate. The whispering frames OSI as “gatekeepers” and “an opinion” without acknowledging that a license which fails the open source definition (OSD) does so because users of the software will be unclear they have all the rights they need to proceed, so have to gain clarification and permission, chilling collaboration.

Lobbyists request arrangements where deployed copies must be counted, or royalties must be negotiated (on a “fair and reasonable” basis of course), or where the application field must be analysed and classified to check deployment is permitted. They have also argued for other scenarios against which the OSD protects. The only developers that show up in projects deploying these measures will be the ones whose employers have already negotiated in private. The whole point of promoting open source is the network effect it catalyses. These measures poison the catalyst.

We need to make sure that the voices representing the open source movement at large explain to the Commission and others that the open source effect they seek cannot be generated by accommodating companies that have already exploited and are now discarding that effect, retaining only its name for marketing purposes. Nor can the open source effect be enjoyed in contexts where patent royalties or market control are so important that developers can’t be allowed to be the technology kingmakers. Paving the ground to protect certain trees will stop the woodland growing.

The open source effect demands software freedom and “pragmatically” curtailing it simply prevents the success of “open initiatives” by chilling the open source effect on which they rely. The business models of the companies involved — and indeed the interpretation of a license during a legal confrontation between companies — are only a fraction of the full story.

We should be advocating for faithfulness to community enablement through software freedom, and the cultivation of the meshed society, not fixating on the strategy adjustments of those who would successfully exploit (or smother) the open source effect without embracing it. That message needs loudly repeating in each government innovation session, each SDO open source strategy think-tank. Woods, not just trees.